ns 
fe 4 iy fom alcohol, 50 cc. of Cher being ea he first time 
ment afterwards. 
to 50 or 60°, in otdet to separate the vaniltie from the small particles (e) 
Ree ea eS 
The ethereal solution is dried over calcium | cn 
pours it into a measuring flask of 100 cc., rinses, fills up to the mark and f en 
5 cc. of the filtrate, in a measuring flask, 5 cc. of a 1 per cent. solution of 
alcohol in 95 per cent. ethyl alcohol are added and afterwards 20 eG; of cones 
lized vanillin, areal dried until the weight remained constant The colot: 
resulting are compared in a colorimeter, it being admissible to dilute the so 
with dilute sulphuric acid (1 he _ concentrated sulphuric acid and 1 vol. of wai 
The vanilla content amounts to = a corresponding to the number of mgs. of vanil 
found in the final solution and g to the weight, in grams, of the vanilla used. The 
value a is obtained with the aid of tables accompanying the treatise. = 
The vanillin content thus determined for Bourbon vanilla oscillated between 23 
and 3.03 per cent., an inferior kind only containing 1.54 per cent. Een 
In case the sanilla is suspected to have been fraudulently covered with vanillin, 
it is advisable to test the outer and inner layers separately, when normal vanilla will 
not show any great differences. : : 
It remains to be seen in how far the methods described will be adopted in analytica 
practice. As far as we are concerned, we hold that a colour reaction will always be 
especially for quantitative estimations, a mere makeshift, which one is glad to dispen 
with if there are more exact tests available. This is the case for either cinnamaldehy 
and vanillin, for the quantitative estimation of which J. Hanus published some time 
ago very useful and reliable methods, according to which cinnamaldehyde’) is broug it 
into reaction with semioxamazide and vanillin?) with #-naphthylhydrazine or, bette 
still, with p-bromophenylhydrazine. These methods are specially suitable for the 
quantitative determination of small quantities of the aldehydes in question, such 
are present in the respective drugs, and have at least the advantage over Fellenberg’ Ss 
tests that they are exactly quantitative estimations in the literal meaning of the » 
the results of which do not depend upon the subjective impression of the. indivi 
observer. : 
We consider recommendable, however, the way of sampling the vanilla and 
of extracting, previous to the distillation of the cinnamon, the cinnamaldehyde 
determination. As to the distillation itself, we piel it with ater =< 
steam. > | Ses es 
in various cinnamon aS Ceylon cinnamon 1.74 to 2.19 per sone: tae 
1.42 per cent., cassia bark 2.08 to 3.81 per cent., flores cassiw 3.70 to 6 per 
1) Zeitschr. Untersuch. d. Nahrungs- u. Genussm. 6 (1903), 817. 
— 4) Ibidem 8 1900) 
7 (1904), 671. ‘ 
